Appropriate statistical models will be developed for the analysis of family data to elucidate the action of major genes governing susceptibility to cancer and similar diseases with variable age of onset; these will include segregation and linkage models, that allow for familial environmental components, for both qualitative and quantitative traits. Computer programs based on these models will be written, documented for general use, and used in simulation experiments to assess the power, efficiency and robustness of the methods of analysis developed. These models and programs will also be used, collaboratively, in the analysis of cancer family data collected by other workers. This will result in the development of appropriate methods, based on the results obtained, for genetic counseling of individuals in cancer families; and for the early detection of cancer-prone individuals to help research in the biochemical pathways involved.